Font Size: a A A

Design Of Identification System Based On Palm Vein

Posted on:2020-03-22Degree:MasterType:Thesis
Country:ChinaCandidate:Y N LiuFull Text:PDF
GTID:2428330602497132Subject:Control Engineering
Abstract/Summary:PDF Full Text Request
With the development of information technology,people make higher demands on information safety.Accurate identification of a person's authentication has become a widely concerned problem.Compared to traditional identity authentication method such as access card and password,biometric recognition technology has been highly valued because of it's good security and high accuracy.Among them,palm vein recognition technology has many advantages,such as good stability,high recognition accuracy,high anti-counterfeiting and easy acceptance by users.In this paper,we mainly study the algorithm of palm vein recognition.The main work and results of the research of this paper are as follows :(1)In this paper,we study and do comparative experiment on ROI extraction method,image enhancement algorithm,image denoising algorithm of palm vein image.At last,we select the ROI extraction method based on the palm rectangle,gray scale transformation image enhance method of CLAHE,image denoising method of mean filtering.So that follow-up research can achieve the best results.(2)In this paper,we study palm vein recognition methods based on NBP and SURF,and make improvement on NBP feature extraction method to improve the accuracy of NBP algorithm.Then we do comparative analysis on two algorithms by experiments.It provides ideas and foundation to subsequent fusion algorithm.(3)In this paper,we analysis the feasibility of fusing NBP and SURF: NBP algorithm is easier and real-time,it can make up for the disadvantage of SURF on recognition time: SURF has the feature of local invariance,it can make up for the disadvantage of NBP when NBP algorithm meets larger displacement and has lower accuracy.Theoretically,the fusion of two algorithms can achieve better results.And then,we put forward a algorithm fused texture feature and locally invariant feature to improve the accuracy and shorten the time of palm vein recognition.At last,we test the algorithm by experiment.
Keywords/Search Tags:palm vein recognition, NBP features, SURF features, fusion recognition
PDF Full Text Request
Related items